Free Preprint: Model-based Methods for Continuous and Discrete Global Optimization

The paper “Model-based Methods for Continuous and Discrete Global Optimization” (T. Bartz-Beielstein, M. Zaefferer) is available for download via Cologne Open Science:
Surrogate models gained importance for discrete optimization problems. This paper presents a survey of model-based methods, focusing on continuous optimization. It introduces a taxonomy, which is useful as a guideline for selecting adequate model-based optimization tools, and provides details for the case of discrete optimization problems. Six strategies for dealing with discrete data structures are introduced.
A new approach for combining surrogate information via stacking is proposed. The implementation of this approach will be available in the open source R package SPOT2. The article concludes with a discussion of recent developments and challenges in both application domains.

Address = {TH K{\”o}ln},
Author = {Thomas Bartz-Beielstein and Martin Zaefferer},
Institution = {Fakult{\”a}t f{\”u}r Informatik und Ingenieurswissenschaften (F10)},
Month = {August},
Number = {8/2016},
Title = {Model-based Methods for Continuous and Discrete Global Optimization},
Type = {Schriftenreihe CIplus},
Year = {2016}}